HAL Id: hal-02318291https://hal.archives-ouvertes.fr/hal-02318291

Submitted on 12 Oct 2020

HAL is a multi-disciplinary open accessarchive for the deposit and dissemination of sci-entific research documents, whether they are pub-lished or not. The documents may come fromteaching and research institutions in France orabroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, estdestinée au dépôt et à la diffusion de documentsscientifiques de niveau recherche, publiés ou non,émanant des établissements d’enseignement et derecherche français ou étrangers, des laboratoirespublics ou privés.

MOKE Magnetometer Studies of Evaporated Ni andNi/Cu Thin Films onto Different Substrates

L. Kerkache, A. Layadi, M. Hemmous, A. Guittoum, M. Mebarki, NicolasTiercelin, A. Klimov, Vladimir Preobrazhensky, Philippe Pernod

To cite this version:L. Kerkache, A. Layadi, M. Hemmous, A. Guittoum, M. Mebarki, et al.. MOKE MagnetometerStudies of Evaporated Ni and Ni/Cu Thin Films onto Different Substrates. SPIN, World ScientificPublishing, 2019, 09 (01), pp.1950006. �10.1142/S2010324719500061�. �hal-02318291�

1

MOKE magnetometer studies of evaporated Ni and Ni/Cu thin films onto

different substrates

L. Kerkache (1) , A. Layadi (1)*, M. Hemmous (2), A. Guittoum (2), M. Mebarki (3), N.

Tiercelin (4), A. Klimov (4), V. Preobrazhensky (4) and P. Pernod (4)

(1) L.E.S.I.M.S., Département de Physique, Université de Sétif 1, Sétif 19000, Algeria

(2) Nuclear Research Centre of Algiers, 2 Bd Frantz Fanon, BP 399, Alger-Gare, Algiers, Algeria

(3) Centre de Recherche en Technologie des Semi-conducteurs pour l’Energétique (CRTSE),

2, Bd Frantz Fanon, BP 140 Alger 7- Merveilles 16038, Algeria

(4) Univ. Lille, CNRS, Centrale Lille, ISEN, Univ. Valenciennes, UMR 8520 IEMN, LIA

LICS/LEMAC, F-59000 Lille, France

*Corresponding author: alayadi@univ-setif.dz or A_Layadi@yahoo.fr (A. Layadi)

Abstract

The Magneto-Optic Kerr Effect (MOKE) technique has been used to investigate the

magnetic properties of Ni thin films, with thickness t ranging from 9 to 163 nm, evaporated

onto several substrates (glass, Si (111), mica and Cu) with and without an evaporated Cu

underlayer. The MOKE observations were correlated with the surface morphology inferred

from Scanning Electron Microscope images and with the structural properties (grain size and

strain). Some interesting behaviors of the coercive field (with values in the 2 to 151 Oe

range), the squareness (between 0.1 and 0.91) and the saturation field (25 to 320 Oe) are

observed as a function of t, the substrate and the Cu underlayer. A thickness dependent stress-

induced anisotropy is found in these films. The differences between the present MOKE results

and the ones obtained from the Vibrating Sample magnetometer (VSM) are highlighted. The

former describe the surface magnetism of these systems, while the latter are attributed to the

whole volume of the sample.

Keywords: Ni films, Ni/Cu films, Evaporation, MOKE, SEM

2

1. Introduction

A lot of important phenomena are observed in ferromagnetic thin films and

multilayers, including Ni films and Ni based multilayers; theses features are interesting for

both fundamental research and technological applications [1-5]. The physical properties of

thin films strongly depend on the method and the conditions of deposition, on the substrate,

and the film thickness. Several works have been reported on Ni films [6-12] and on the

Ni/Cu/substrates bilayers [13, 14]. Among the effect of Ni thickness on the physical

properties, one may cite (i) the transition from an out-of plane to an in-plane magnetization

easy axis, (ii) the effect of the lattice strain on the magnetic moment reduction and (iii) the

change from a fine grain structure to a column structure [15-18].

We have used the thermal evaporation to grow series of Ni thin films in the 9 to 163

nm thickness range, deposited on several substrates (glass, Si (111), mica and Cu) with and

without an evaporated Cu underlayer. The structural properties of these systems have been

published elsewhere [19].Also, we have studied some magnetic properties of these systems as

inferred from a Vibrating Sample Magnetometer (VSM) set-up [20].

In the present work, we used the Magneto-Optic Kerr Effect (MOKE) technique to

investigate the magnetic properties as a function of the substrates, the Ni thickness and a Cu

underlayer. The MOKE technique has been used for a long time as a means for the magnetic

characterization of materials [20-22]. The magneto-optical (MO) properties of Ni have been

studied by MOKE measurements [23, 24]. Here, in our investigation, we focused on the

effect of the substrates, the Ni thickness and the Cu underlayer on the magnetic properties of

Ni thin films, such as the coercive and saturation fields, the squarness and the magnetic

anisotropy of Ni/substrates films and Ni/Cu/substrates bilayer films. Also, we attempted to

correlate the structural and the magnetic properties of Ni and Ni/Cu films, through results

from Scanning Electron Microscopy (SEM) and X-ray diffraction experiments. The

3

deposition conditions and the characterization techniques are shown in section 2. In section 3,

we show the results pertaining to the magneto-optical (MOKE) hysteresis curves along with

the SEM images. We discuss the effect of thickness, grain size, Cu underlayer and substrates

on the magnetic parameters of the hysteresis curves. A summary and a conclusion are given in

section 4.

2. Experimental details

The Ni thin films were deposited onto glass, Si (111), mica and Cu substrates using

thermal evaporation method from a 99.995 % purified Ni powder. For the Ni/substrates, the

pressure was approximately 7.69 x 10-7 mbar before evaporation; it rises to 2.86 x 10-6 mbar

during the evaporation. For the Ni/Cu/substrates, the same purity for the powder (Cu) was

used to evaporate the Cu underlayer. The pressure ranges from 1.52 x 10-6 and 1.72 x 10-6

mbar before evaporation to 1.58 x 10-5 and 1.74 x 10-5 mbar during the evaporation for the

deposition of the Ni and Cu films respectively. The Ni thickness t ranges from 9 to 163 nm

and 8 to 67 nm for the Ni/substrates and Ni/Cu/substrates systems, respectively. For the

Ni/Cu/substrates system, the Cu underlayer has two thicknesses, 90 nm (with t = 24 nm) and

52 nm (t = 8, 14 and 67 nm). Details on the preparation conditions and on the structural

properties of these systems have been reported elsewhere [19]. The surface morphology is

studied by means of a Scanning Electron Microscope (SEM). The magneto optical Kerr effect

(MOKE) technique has been used to describe the magnetic properties of these samples. Kerr

effect experiments were done to get the hysteresis curves, Kerr rotation θK vs applied field H.

The measurements were done, at room temperature, in the longitudinal configuration with a

magnetic field H varying from - 0.5 to 0.5 kOe.

3. Results and discussion

4

In Fig. 1, we show examples of the magneto-optic Kerr effect (MOKE) hysteresis

loops, for the same Ni thickness (t = 50 nm) and different substrates (glass, Si, mica and Cu)

(Fig. 1a). In Fig. 1b, we display the curves pertaining to the Ni/Cu/substrate system with t =

24 nm and tCu= 90 nm. Fig.1a is meant to show the effect of the substrate while by comparing

Fig.1b and Fig.1a, we observe the effect of the Cu underlayer.

The effect of the substrate on the hysteresis curves can be clearly seen by comparing

the four figures in Fig. 1a. One can see the differences in the curve shapes, the curve area and

the different parameters. For the shape for instance, the Ni/glass curve is square while that of

the Ni/Cu curve is a more hard-axis like one, this is due to a change in magnetic anisotropy as

will be discussed later on. For the area, Ni/glass has the smallest area compared to the other

curves; the area under the hysteresis curve is generally related to the energy needed for the

motion of the Bloch walls and the reorientation of the magnetic moments of the domains [25].

The magnetic parameters characterizing the curves such as the coercive field, the squarness

and the saturation field are also different for each case; theses values have been measured for

all samples and will be discussed in the following. One can also see the difference in the

curves with the Cu underlayer. Moreover, one should distinguish between the Cu as a

substrate which is a bulk material and the Cu underlayer which is an evaporated thin film on

top of the substrate; we will see that the effect on the Ni properties is quite different.

Furthermore, Kerr Effet being very sensitive to the film surface state, we are showing

in Fig. 2, examples of SEM surface images of some samples along with the hysteresis loops

of theses same samples. One can easily note that there are modifications in both the film

surface and the corresponding hysteresis curve; for instance, when the surface is uniform as in

Fig. 2.a, a square shaped curve is observed and for a less uniform surface, a change of the

curve shape is seen, the curves become broader (see Fig.2 b and c). Thus, we believe that at

5

least, part of the differences in the shape of the hysteresis curves may be due to the effect of

the surface morphology and the surface roughness of Ni films.

In the following we will discuss the coercive field, HC, the squarness S and the

saturation field HS. Recall that for each thickness, the four samples were prepared in the same

run, i.e. under the same conditions. Thus, for a given thickness, any changes in the magnetic

properties of the Ni thin film are due to the effect of the substrate in the Ni/substrate system

and to that of the substrate and the Cu underlayer for the Ni/Cu/substrate system. As a general

remark, we note that there are broad ranges of values for HC, S and HS, depending on the

substrate, the Ni thickness and the Cu underlayer. Indeed, HC values vary from 2 to 151 Oe,

the S values can be as low as 0.1 and as high as 0.91 and the saturation field values are found

in the 25 to 320 Oe range. Beyond the interesting physics concerning the origins of the values

and the behaviors of these magnetic parameters, these broad ranges of values can be quite

useful for technological applications, such as in magnetic and magneto-optical recordings; as

these large sets of values may allow one to tune the value of one of these magnetic parameters

to the required one for a given application by choosing the substrate, varying the thickness or

adding or not a Cu underlayer.

In Fig. 3, we show the variation of the coercive field, HC, as a function of the Ni

thickness, t. For the Ni/substrates system (Fig. 3.a), Ni/glass and Ni/Si series have the same

behavior: the coercive field monotonically increases with increasing thickness, from about 2

to 54 Oe; for the Ni/glass samples, there is almost a linear increase. On the other hand, the

HC vs t curves for Ni/mica and Ni/Cu go through a minimum at t = 50 nm; the HC values at

this minimum are equal to 32 Oe and 19 Oe for the films deposited on mica and Cu

substrates, respectively. The Ni/glass samples are characterized by the lowest HC values for

6

all thicknesses. For the thickest film (167 nm), the HC values are the same for Ni on glass, Si

and Cu, and slightly higher for Ni on mica.

For the effect of Cu underlayer, i.e., the Ni/Cu/substrates system, see Fig. 3.b. We note

that in the Ni/Cu/mica series, there is a linear increase of the coercive field with increasing Ni

thickness. While for the other substrates, we observe a minimum value at t = 24 nm; it should

be noted that this particular sample has a Cu thickness equal to 90 nm, different from the

other bilayers where the Cu thickness is 52 nm. Thus, this low value might be due to the

relatively higher Cu underlayer thickness; a similar behavior was observed by C. A. F. Vaz et

al. [26], where they found that HC decreases with increasing Cu thickness. Note also that for

the smallest thickness (08 nm), the bilayers (Ni/Cu) deposited on glass, Si (111) and mica

have almost the same coercive field value (2 Oe), i.e. the coercive field is independent of the

substrate; also for this thinner film and at least for Ni/Cu/glass and Ni/Cu/Si, the coercive

field remains low and the same (2 Oe) as the ones measured in Ni/glass and Ni/Si, i.e, for the

thinner film, HC is low, independent of the substrate and is not affected by the Cu underlayer .

On the other hand, for the thickest sample (t = 67 nm), Ni/Cu/Si is characterized by a much

higher HC value (151Oe) compared to the other samples which have closer values (about 70

Oe).

For the sake of comparison with reported Ni thin film HC values, we can cite few

works in this matter. The coercive field values found here for the Ni/substrate system are

comparable to those reported in refs. [27] and [28], but are lower than those reported in ref.

[29]. The variation of HC with thickness of the Ni/Cu/Si is similar to the one observed by H.

Hwang and al. [30]; also in this same system, the HC values are similar to those found by G.

Gubbiotti and al.[31], but larger than those measured by B. Schirmer and al. [32]. For

7

comparable thickness, A. Sharma et al. [33] found a coercive field of 130 Oe for Ni (40

nm)/Si sample; this value is higher than the one we found (13 Oe for Ni (42 nm)/Si).

For the origins of the HC values and behaviors, the following remarks can be made. As

for the variation of HC with t and the fact that HC is quite high for thicker films, we believe

the reason may arise from magnetic anisotropy; we will come back to this point when

addressing the magnetic anisotropy later on. To get more insights on the origins of HC values,

we have studied the variation of HC with the grain size. For the Ni films deposited onto glass

and Si (111) substrates, the coercive field HC values increase with the increase of the average

grain size D. In Table 1, we show examples of such a variation of HC with D. For the

Ni/mica, the HC vs D curve has a minimum, but there is a general increase of HC with D;

while no clear variation of HC with D is seen for Ni/Cu. For the Ni/Cu/substrate system,

beyond a grain size value of about 6 to 8 nm, HC increases with grain size for all substrates.

For Ni/Cu/glass and excluding the sample with the larger Cu thickness (90nm), HC increases

with increasing average grain size. For the low D values and for Ni/Cu/Si (111), we note a

monotonous decrease of HC with D (see Table 1), then an increase for the high D value; no

clear variation is noticed for the other substrates in the low D values, but HC increases for

higher D. From this kind of variation of HC vs D, one may conclude the following. First, one

can rule out the pinning of domain walls at the grain boundaries as responsible for the HC

behavior; indeed, if pinning at the grain boundaries were the origin of the coercive field

values, HC would decrease with increasing D. We observe the inverse behavior, thus pinning

at the grain boundaries is not the dominant factor in the values of the coercive field in our

case. Second, this kind of variation, i.e the increase of HC with D, was observed in many

8

materials; for instance, M. Mebarki et al.[34] found this variation in Fe/glass thin film for low

D values; this behavior is well described by some models such as in the Hoffmann’s

theory[35]. Finally as mentioned above about the SEM images, where different magnetic

behaviors are correlated with different surface morphologies, the surface state (the rugosity,

the size, shape and distribution of the grains) may play an important role in the high values of

HC measured in some samples, it may lead to a lot of pinning sites [29] which constitute a

barrier for the motion of the magnetic domain walls, increasing thus the coercive field.

Therefore, in our study, we believe that the main origins of the coercive field behavior are the

magnetic anisotropy, the grain size effect and the surface morphology which depend on the

substrate, the Ni thickness and the Cu underlayer.

The squarness S is defined as the ratio of the remnant Kerr rotation θKR to the

saturation Kerr rotation θKS, i.e. S = θKR/θKS. In general, the squareness gives information on

the remnant state of a magnetic material which is important in fundamental research and also

useful in technological applications. The squareness value may be associated to the

mechanisms of the magnetization reversal. A small S value may indicate a magnetization

rotation; on the other hand, high S values may be associated with domain wall nucleation and

motion [33, 36, 37].

For to the Ni/substrates system, the squarness values are shown in Fig.4.a. We found a

whole range of S values, from as low as 0.1 to as high as 0.91. The highest S values are

observed for a Ni thickness of about 42 to 50 nm in Ni/glass (0.91), Ni/Si (0.91), Ni/mica

(0.89) and Ni/Cu (0.87). Note for comparison that for polycrystalline bulk Ni material, S is

equal to 0.866 [38]. In the low thickness range, we note an increase of S for Ni/glass and

Ni/Si when t increases from 9 nm to 42 nm (for Si) and to 50nm (glass). Beyond this

thickness (50 nm), a monotonous decrease of S is seen for all substrates. For the thickest film,

the highest squareness is noted in the Ni/mica sample (S = 0.67), while the lowest S value (S

9

= 0.45) is observed in Ni/Cu and Ni/Si. The latter one (S = 0.45) is comparable to that found

by K. Zhang and al. [38] for Ni/Si films. A. Sharma et al. [33] found a value of S equal to

0.82 for Ni (40 nm)/Si; this value is slightly lower than the one we found (S = 0.91) for Ni(42

nm)/Si.

For the Ni/Cu/substrate system, see Fig. 4. b. For the thinner film (8-9 nm), we note

that the Cu underlayer induced a decrease of S for the Ni/glass sample but did not affect the S

value of the Ni/Si one, even though these two samples were prepared in the same run,

compare Fig. 4a and 4b; the decrease in Ni/Cu/glass is quite large (from 0.72 for Ni/glass to

0.1 for Ni/Cu/glass). The variation of S with the Ni thickness does depend on the substrate.

For Ni/Cu/glass, S increases in a monotonous manner with t from 0.1 at t = 9 nm to 0.77 at t

= 67 nm. The opposite is seen for Ni/Cu/Cu where S gradually decreases with increasing

thickness. For the other substrates (Si and mica), S increases with t up to t = 24 nm (and tCu =

90nm), then slightly decreases when t is increased further. For the largest thickness (t = 67

nm), S is about the same for Ni/Cu on Si, mica and Cu and slightly higher for Ni/Cu on glass.

We also looked at the variation of S with grain size D. For the Ni/substrate system, in

the low D values, S increases with increasing D up to D = 5.5 ± 0.2 nm and then drops for

higher D values, see Table 1 for Ni/glass and Ni/Si; this is true also for Ni/Cu but not for

Ni/mica, where S decreases with D. For the Ni/Cu/substrates system, a monotonous increase

of S with D is seen in Ni/Cu/glass. The opposite occurs for Ni/Cu/Cu where S decreases with

increasing D. While no clear variation is seen for the Si and mica substrates.

For the Ni/substrate system, the variation of the saturation field HS with Ni film

thickness t is shown in Fig. 5a. We can see that regardless of the substrate, the saturation

magnetic field monotonously increases with increasing thickness t. The saturation field values

for Ni/Cu thin films are the largest for all thicknesses, while those corresponding to the

Ni/mica are the lowest. For comparison with other work in the same conditions, we may cite

10

that A. Sharma et al.[33] found a saturation field of 230 Oe for Ni (40 nm)/Si; this value is

higher than our result (98 Oe) for Ni(42 nm)/Si.

For the Ni/Cu/substrate system, the saturation field vs Ni thickness is shown in Fig.

5b. Here also, we observe an overall increase of HS with thickness for all substrates; the Cu

underlayer did not affect this HS behavior with t. However, there is a monotonous increase

(an almost linear increase with t) for Ni/Cu on Cu and mica substrates, while for Ni/Cu on

glass and Si substrates; there is a thickness range (14 to 24 nm) where HS seems to remain

constant before increasing again. For the thinner samples (8 nm), Ni/Cu deposited on glass, Si

and mica have the same HS values. For the thicker samples (67 nm), the HS values for Ni/Cu

on Si, mica and Cu substrates are the same and the value corresponding to Ni/Cu/glass is

somewhat higher. We can see that the saturation field strongly depends on the substrates, the

Ni thickness and the Cu underlayer. We have also studied the variation of the saturation field

HS with grain size D. For the Ni/glass, Ni/Si and Ni/mica samples, we observe a monotonous

increase of HS with D, see Table 1 for Ni/glass and Ni/Si, as examples. On the other hand, for

Ni films deposited on Cu substrate, HS has a monotonous decrease with increasing D. For the

effect of Cu underlayer, we observed an overall increase in the saturation field with the

increase of D for the Ni/Cu/glass (see Table 1) and Ni/Cu/Cu. However, for Ni/Cu on Si and

mica, we observed a decrease of HS with D in the low D values (up to D = 7 nm) then HS

increases with D as D is further increased.

It is worth noting that the saturation field values are generally associated with the

magnetic anisotropy. For a given direction, a low (high) saturation field value may indicate

that this direction is an easy (hard) direction for the magnetization. In our case, the saturation

field values are much lower in the thinner films than in the thicker ones (see Fig. 5). We may

11

infer that a magnetic anisotropy exists; it is strong in thinner films and diminishes as the film

thickness increases; it makes the plane direction a much easier one in thinner films than in

thicker ones. We conclude this from the HS behavior and also from the shape of the hysteresis

curves; the thinner films curves are characterized by a square, easy direction-like curves with

low HS and HC, while as the Ni thickness increases, the hysteresis curves become more and

more hard direction-like curves with high HS and HC values. In order to investigate the origin

of this magnetic anisotropy, we made a correlation between the structural and magnetic

properties. In these samples, we found that the strain ε values in thinner films are relatively

higher than those measured for thicker ones; thinner films are subjected to a higher stress, as

the thickness increases, the stress decreases, i.e. the stress is relieved. Thus higher stress

values are associated with lower saturation field values; see Table 1, compare the magnitude

of the strain and the HS values for Ni/glass and Ni/Si, with and without Cu underlayer. We

may conclude from this correlation between ε and HS that the origin of the magnetic

anisotropy is the stress. Thus, a stress-induced magnetic anisotropy is present in these Ni

films; it is stronger in thinner films than in thicker ones. It is detected for Ni on all substrates;

however the magnitude of this anisotropy is different from one substrate to the other. In fact

magneto-elastic effects have been reported to exist in Ni films under some conditions [39].

Finally, it is interesting to compare the present results obtained by the MOKE

technique with the ones measured on these systems by means of the Vibrating Sample

Magnetometer (VSM) method [20]. We noticed, indeed that the characteristics of the MOKE

hystersis curves are different from the VSM ones. This is expected; the effect was seen in

several ferromagnetic materials and was reported by a lot of researchers [40]. One of the main

reasons for the differences is the fact that the MOKE technique, based on the interaction of

light with the sample, concerns only the surface or a small region near the surface (the

12

penetration depth of the light); thus the parameters derived from the MOKE method describe

the surface magnetism of the sample. On the other hand, the parameters obtained from VSM

are average values over the whole volume of the sample. In our case, we can cite few of the

main differences between the results. (i) Starting from the last point we discussed, i.e. the

saturation field HS values and the magnetic anisotropy, we found that the HS values measured

from VSM are between 280 Oe and 2400 Oe for the Ni/substrate system and from 520 to

about 3200 Oe for the Ni/Cu/substrate; these values are much higher than the HS values we

derived here from MOKE which range from 25 to a maximum of 320 Oe for both systems. It

follows that the strength of the magnetic anisotropy over the entire sample volume is more

important than that acting on the magnetization on the surface of the sample. (ii) Moreover,

for the saturation field and magnetic anisotropy, with the MOKE investigation, we found that

there is an overall increase of the HS values as t increases, for all substrates and for both

systems (compare the HS values at t = 8 nm and t = 163 nm in Fig. 5a and the values at t = 9

nm and t = 67 nm in Fig. 5b). When using the VSM, we noticed that again the monotonous

increase of HS with t is seen in the Ni/substrates for all substrates. However for the

Ni/Cu/substrate, the increase of HS with t is followed for the glass and mica substrates, but for

Ni/Cu/Si and Ni/Cu/Cu, we see the opposite, HS decreases with increasing t. This means that

not only the magnitude of the magnetic anisotropy is different in the volume (VSM) and in

the surface (MOKE) but also the easy axes in both cases are different. For the MOKE

technique, the fact that HS increases for the thicker films means that the plane becomes a

relatively harder direction compared to the thinner films and this applies for all substrates. On

the other hand, from VSM, the decrease of HS for the thicker films, for the Si and mica

substrates, means that the plane in these films is becoming a relatively easier direction for the

magnetization. Thus once again, the volume of the sample may consist of regions (including

13

the surface) with different anisotropies; the VSM gives an average value and a resultant easy

axis for the whole volume, while with MOKE we may determine the easy axis and the

strength of the anisotropy for the surface. (iii) The coercive field values derived from MOKE

are lower than those obtained from the VSM for most samples, see Table 2 for some samples

given as examples. This is somewhat logical, since as pointed before, part of the HC values

may arise from magnetic anisotropy and we noted that the anisotropy in the volume is

different from that at the surface. (iv) Finally, we measured different squareness values (see

Table 2) from the two techniques. In most samples, the S values from VSM are lower than

the ones from MOKE. Depending on the interest of a researcher, one or the other method may

be appropriate. For researchers interested in magneto-optical applications or surface effects,

the present MOKE results might be useful and meaningful.

4. Conclusion

The Magneto-Optic Kerr Effect (MOKE) technique is used to probe the surface

magnetism of a series of Ni thin films, with thickness t in the nanometer range, evaporated

onto glass, Si (111), mica and Cu substrates, and also a second series with a Cu underlayer

and the same substrates, i.e Ni/Cu/substrates. We found that the values and the behaviors of

the coercive field HC, the squareness S and the saturation field HS strongly depend on the

substrate, the Ni thickness and the Cu underlayer. For the Ni/glass and Ni/Si series, HC

monotonically increases with increasing t; while for Ni/mica and Ni/Cu, the HC vs t curves

go through a minimum value at t = 50 nm. The Ni/glass samples are characterized by the

lowest HC values. We believe that the main origins of the HC values and behaviors are the

magnetic anisotropy, the grain size effect and the surface morphology. For the latter, different

magnetic behaviors are correlated with different surface morphologies as inferred from SEM

images. The highest S values (0.87 - 0.91) are observed for t around 50 nm. For the thinnest

14

film (8-9 nm), Cu underlayer induced a large decrease of S for the Ni/glass sample (from 0.72

for Ni/glass to 0.1 for Ni/Cu/glass), but did not affect the S value of the Ni/Si one. The

saturation field HS monotonously increases with increasing t, regardless of the substrate; the

Cu underlayer did not affect this HS behavior with t. From the correlation of the HS and the

stress values, we may infer that a stress-induced magnetic anisotropy is present in these Ni

films; it is stronger in thinner films than in thicker ones. Finally, we pointed out the

differences between the MOKE results discussed here and the ones derived from VSM. Such

a comparison allows one to distinguish the surface magnetism described by the MOKE from

the properties of the whole volume associated with the VSM measurements. Among the

differences, we found that the magnetic anisotropy for the volume is more important than that

at the surface.

15

References

[1] P. Grünberg, R. Schreiber, Y. Pang, M. B. Brodsky, and H. Sowers, "Layered Magnetic

Structures: Evidence for Antiferromagnetic Coupling of Fe Layers across Cr Interlayers",

Phys. Rev. Lett.57, 2442 (1986).

[2] Y. Wang, C. H. de Groot, D. Claudio-Gonzalez, and H. Fangohr, "Magnetoresistance in a

lithography defined single constrained domain wall spin-valve", Appl. Phys. Lett.97, 262501

(2010).

[3] H. He, Z. Zhang, B. Ma, and Q. Jin, “[Co/Ni]N-based synthetic antiferromagnet with

perpendicular anisotropy and its application in pseudo spin valves” , IEEE Trans. Magn.46,

1327 (2010).

[4] J. A. C. Bland and B. Heinrich. Ultrathin Magnetic structures I: " An Introduction to the

Electronic, Magnetic and Structural Properties", Springer, 2005.

[5] A. Layadi, J.O. Artman, R.A. Hoffman, C.L. Jensen, D.A. Saunders and B.O. Hall,

"Coupling of Ni and NiFe films through an intervening non-magnetic film", J. Appl. Phys. 67,

9, pp. 4451-4453, (1990).

[6] A. L. Dolgiy, S.V. Redko, I. Komissarov, V.P. Bondarenko, K.I. Yanushkevich, S.L.

Prischepa. "Structural and magnetic properties of Ni nanowires grown in mesoporous silicon

templates ", Thin Solid Films 543 (2013) 133–137.

[7] Z. B. Guo, W.B. Mi, Q. Zhang, B. Zhang, R.O. Aboljadayel, X.X. Zhang, "Anomalous

Hall effect in polycrystalline Ni films", Sol. Sta. Com. 152 (2012) 220–224.

[8] Z. Yang, H. Qiu, B. Hu, G. Zou, " Structural and electrical properties of Ni films sputter-

deposited on HCl-doped polyaniline substrates", Vacuum. 99 (2014) 192-195.

[9] C. Nacereddine, A. Layadi, A. Guittoum, S.-M. Cherif, T. Chauveau, D. Billet, J. Ben

Youssef, A. Bourzami, M. H. Bourahli. , "Structural, electrical and magnetic properties of

evaporated Ni/Cu and Ni/glass thin films", Mater. Sci. Eng. B 136 (2007) 197–202.

16

[10] S.-M. Chérif, A. Layadi, J. Ben Youssef, C. Nacereddine, Y. Roussigné, "Study of the

magnetic anisotropy in Ni/Cu and Ni/glass thin films", Physica B 387 (2007) 281–286.

[11] M. Hemmous, A. Layadi, A. Guittoum, A. Bourzami, A. Benabbas, "Effect of deposition

rate and thickness on the structural and electrical properties of evaporated Ni/glass and

Ni/Si(100) thin films", Microelectronics. J. 39 (2008) 1545 – 154 9.

[12] 48. B. Ghebouli, A. Layadi and L. Kerkache, "Effect of the substrate on the structural

and electrical properties of dc sputtered Ni thin films.", Euro. Phys. J.: AP 3, 35-39, (1998).

[13] R. Naik, M. Ahmad, G. L. Dunifer, C. Kota, A. Poli, Ke Fang, U. Rao and J. S. Payson,

"Growth and magnetic studies of epitaxial films of Fe, Co and Ni on Cu(100)/Si(100)", J.

Magn. Magn. Mater. 121 (1993) 60-64. 

[14] H. J. Hug, B. Stiefel, A. Moser, I. Parashikov, A. Klicznik, D. Lipp, H.-J. Güntherodt, G.

Bochi, D. I. Paul, R. C. O’Handley, "Magnetic domain structure in ultrathin

Cu/Ni/Cu/Si(001) films (invited)" , J. Appl. Phys.79, 5609 (1996).

[15] J. S. Lee, Y. J. Park, J. H. Song, K. H. Chae, J. Lee, C. N. Whang, K . Jeong, D. H. Kim,

S. C. Shin, "Modification of interface magnetic anisotropy by ion irradiation on epitaxial

Cu/Ni/Cu(002)/ Si(100) films", Phys. Rev. B 69. 172405 (2004).

[16] J. Lee, G. Lauhoff, M. Tselepi, S. Chope, P. Rosenbusch, J. A. C. Bland, H. A. Dürr, V.

D. Laan, J. Ph. Schillé, J. A. D. Matthew, "Evidence for a strain-induced variation of the

magnetic moment in epitaxial Cu/Ni/Cu/Si(100) structures", Phys. Rev. B 55. 15103 (1997).

[17] G. Lauhoff, J. Lee, J. A. C. Bland, J. Ph. Schill, G. van der Laan, "Ferromagnetic

coupling and anisotropy in epitaxial Cu/Co/Ni/Cu (0 0 1)", J. Magn. Magn. Mater. 177-181

(1998) 1253-1254.

[18] G. B. Cho, K. K. Cho, K. W. Kim, "Effects of Ni film thickness on the structural

stability of Si/Ni/Cu film electrodes", Mater. Lett. 60 (2006) 90 – 93.

[19] M. Hemmous, A. Layadi, A. Guittoum, N. Souami, M. Mebarki and N. Menni.

“Structure, surface morphology and electrical properties of evaporated Ni thin films: Effect of

substrates, thickness and Cu underlayer”, Thin Solid Films 562 (2014) 229–238.

17

[20] M. Hemmous, A. Layadi, L. Kerkache, N. Tiercelin, V. Preobrazhensky and P. Pernod,

“Magnetic properties of evaporated Ni thin films: Effect of substrates, thickness and Cu

underlayer ”, Metallurgical and Materials Transactions A, Vol. 46, issue 9 (2015), 4143-4149.

[20] C. C. Robinson, "Longitudinal Kerr Magneto-Optic Effect in Thin Films of Iron, Nickel,

and Permalloy", J. Opt. Soc. Am.53, 681 (1963).

[21] S. Visnovsky, V. Parızek, M. Nyvlt, P. Kielar, V. Prosser, and R. Krishnan, "Magneto-

optical Kerr spectra of nickel", J. Magn. Magn. Mater.127, 135 (1993).

[22] A. Hubert and R. Schäfer, "Magnetic Domains: The Analysis of Magnetic

Microstructures", Springer-Verlag, Heidelberg, 2000.

[23] J. L. Erskine and E. A. Stern, "Magneto-optic Kerr Effect in Ni, Co, and Fe", Phys. Rev.

Lett.30, 1329 (1973).

[24] K. Nakajima, H. Sawada, T. Katayama, and T. Miyazaki, "Effects of the surface and

interface on the magneto-optical properties in (Co, Ni)/Cu(001) ultrathin films", Phys. Rev.

B54, 15950 (1996).

[25] J. P. Bailon et J. M. Dorlot, "Des Matériaux", Presse Internationale Polytechnique, third

edition, Canada, 2000.

[26] C. A. F. Vaz and J. A. C. Bland. "Dependence of the coercive field on the Cu overlayer

thickness in thin Co/Cu(001) and Ni/Cu(001) fcc epitaxial films", J. Appl. Phys. 89, 7374

(2001).

[27] S. Z. Wu, G. J. Mankey, F. Huang, R. F. Willis, "Spin reorientation transition in Ni films

on Cu(100)", J. Appl. Phys. 76. 6434 (1994).

[28] E. E. Shalyguina, M. A. Mukasheva, N. M. Abrosimova, L. Kozlovskii, E. Tamanis, A.

N. Shalygin "The influence of annealing on magnetic and magneto-optical properties of iron

and nickel films", J. Magn. Magn. Mater. 300 (2006) e367 – e370.

[29] J. Pflaum, E. Hubner, Th. Zeidler, T. Schmitte, J. Pelzl, "Studies of the influence of the

local growth characteristics on the magnetic properties of electrolytically deposited Ni films",

Thin Solid Films 318 (1998) 186–188.

18

[30] H. M. Hwang, J. C. Park, D. G. You, H. S. Park, K. Jeong, and J. Lee, T. G. Kim and J.

H. Song, " Spin-reorientation transition of epitaxial Cu/Ni/Cu (001) structure", J. Appl. Phys.

93, 7625 (2003).

[31] G. Gubbiotti, G. Carlotti, F. D'Orazio, F. Lucari, R. Bernardini, M. De Crescenzi,

"Magnetic anisotropy in epitaxial Cu/Ni/Cu/Si(1 1 1) ultrathin films studied by magneto-

optical Kerr effect and Brillouin spectroscopy", J. Magn. Magn. Mater. 177-181 (1998)

1259-1261.

[32] B. Schirmer and M. Wuttig, "Antiferromagnetic coupling in fcc Fe overlayers on

Ni/Cu(100)", Phys. Rev. B 60 (1999) 12945 (5).

[33] A. Sharma, S. Tripathi, K.C. Ugochukwu, J. Tripathi, "Magnetic and structural properties

of Ni nanocaps deposited onto self assembled nanosphere array", Thin Solid Films 536 (2013)

249 –255.

[34] M. Mebarki, A. Layadi, L. Kerkache, N. Tiercelin, V. Preobrazhensky, and P. Pernod,

“Surface morphology and magnetic properties of evaporated Fe/Si and Fe/glass thin films”,

Applied Physics A, Vol. 120, issue 1 (2015) 97-104.

[35] H.S. Jung, W.D. Doyle and S. Matsunuma, "Influence of underlayers on the soft

properties of high magnetization FeCo films", J. Appl. Phys. 93, 6462 (2003).

[36] J. Swerts, K. Temst, N. Vandamme, C. Van Haesendonck, Y. Bruynseraede, "Interplay

between surface roughness and magnetic properties in Ag/Fe bilayers", J. Magn. Magn.

Mater. 240 (2002) 380.

[37] L. Nzoghé-Mendome , A. Aloufy, J. Ebothé, "Nanoscale Topography and Magnetic

Structure of Nanocrystallized Nickel Electrodeposits", Mol. Cryst. Liq. Cryst. 555 (2012) 32.

[38] K Zhang, K P Lieb, N Bibic, N Pilet, T V Ashworth, M A Marioni and H J Hug, "

Microstructural and magnetic properties of thermally mixed Ni/Si bilayers", J. Phys. D:

Appl. Phys. 41 (2008) 095003.

[39] K. Ha, R.C. O’Handley, "Magnetization canting in epitaxial Cu/Ni/Cu/Si(001) films", J.

Appl. Phys. 87 (2000) 5944.

[40] A. Hendrych, O. Zivotsky, J. Jiraskova and I. Matko, "The Surface and Bulk Magnetic

Properties of Fe-Al Alloys ", Acta physica Polonica A, 126 (2014) 58.

19

Table and Figure captions

Table. 1. Strain ε , average grain size D, coercive field HC, squareness S and saturation field

HS for Ni thin films evaporated on glass and Si substrates with and without Cu underlayer.

Table. 2. Comparison between the MOKE and the VSM results, for Ni/substrates and

Ni/Cu/substrates (substrates as indicated); t : the Ni thickness, HC : the coercive field, S :

the squareness and HS : the saturation field.

Fig. 1. The longitudinal Kerr hysteresis loops for (a) Ni/glass, Ni/Si, Ni/Cu and Ni/mica with

t = 50 nm (effect of the substrat); (b) Ni/Cu/glass, Ni/Cu/Si, Ni/Cu/Cu and Ni/Cu/mica with t

= 24 nm and tCu= 90 nm (effect of the Cu underlayer and susbtrate).

Fig. 2. Examples of SEM surface images and the corresponding hysteresis loops for (a) Ni

(42 nm)/Si (111), (b) Ni (14 nm)/Cu (52 nm)/Si (111) and (c) Ni (67 nm)/Cu(52 nm)/Si (111).

Fig. 3. Coercive field HC vs Ni thickness for (a) Ni films on glass, Si, mica and Cu substrates

and (b) Ni/Cu/Substrates with substrates as indicated.

Fig. 4. Squareness S vs Ni thickness for (a) Ni films on glass, Si, mica and Cu substrates and

(b) Ni/Cu/Substrates with substrates as indicated.

Fig. 5. Saturation field HS vs Ni thickness for (a) Ni films on glass, Si, mica and Cu

substrates and (b) Ni/Cu/Substrates with substrates as indicated.

20

Table. 1

Table. 1. Strain ε , average grain size D, coercive field HC, squareness S and saturation field

HS for Ni thin films evaporated on glass and Si substrates with and without Cu underlayer.

Ni/substrates Ni / Cu/ substrates

substrate ε (%) D

(nm)

Hc

(Oe)

S Hs

(Oe)

ε (%) D

(nm)

Hc

(Oe)

S Hs

(Oe)

0.77 4.4 2 0.72 50 -1.72 5.6 02 0.1 32

-0.66 4.9 13 0.86 75 -0.66 6.6 57 0.48 115

-0.22 5.3 15 0.91 100 -0.22 8 10 0.50 110

glass

-0.34 13.7 54 0.51 130 -0.34 14.1 77 0.77 320

0.55 5 2 0.63 62 -0.77 5.6 41 0.75 85

0.20 5.5 13 0.91 98 -0.34 6.2 18 0.81 75

0.02 6.8 26 0.85 112 -1.05 6.6 01 0.60 25

Si (111)

-0.05 15 55 0.45 162 -0.05 13 151 0.65 220

21

Table. 2

Table. 2. Comparison between the MOKE and the VSM results, for Ni/substrates and

Ni/Cu/substrates (substrates as indicated); t : the Ni thickness, HC : the coercive field, S :

the squareness and HS : the saturation field.

MOKE VSM

Substrate t

(nm)

Hc

(Oe)

S Hs

(Oe)

Hc

(Oe)

S Hs

(Oe)

Si 9 2 0.63 62 3.6 0.46 780

glass 163 54 0.51 130 138 0.26 2250

Si 163 55 0.45 160 153 0.23 2400

Ni/substrate

Mica 163 65 0.67 120 115 0.36 1480

glass 8 2 0.10 32 5.7 0.22 1500

glass 67 77 0.77 320 230 0.48 2500

Cu 67 65 0.63 225 57 0.67 1270

Ni/Cu/Substrate

Mica 67 77 0.64 220 168 0.46 1480

22

0 .0

0 .0

0 .0

-2 0 0 0 2 0 0

0 .0

N i/g lass

N i/S i (1 1 1 )

N i/m ica

A p lied m ag n e tic fie ld (O e)

Nor

mal

ized

Ker

r rot

atio

n (θ

K/ θ

KS)

N i/C u

(a )

0 .0

0 .0

0 .0

-2 0 0 0 2 0 0

0 .0

N i/C u /g la ss

N i/C u /S i (1 1 1 )

N i/C u /m ica

Nor

mal

ized

Ker

r rot

atio

n (θ

K/ θ

KS)

N i/C u /C u

(b )

A p lied m ag n e tic fie ld (O e)

23

Fig. 1. The longitudinal Kerr hysteresis loops for (a) Ni/glass, Ni/Si, Ni/Cu and Ni/mica with

t = 50 nm (effect of the substrat); (b) Ni/Cu/glass, Ni/Cu/Si, Ni/Cu/Cu and Ni/Cu/mica with t

= 24 nm and tCu= 90 nm (effect of the Cu underlayer and substrate).

24

Fig. 2. Examples of SEM surface images and the corresponding hysteresis loops for (a) Ni

(42 nm)/Si (111), (b) Ni (14 nm)/Cu (52 nm)/Si (111) and (c) Ni (67 nm)/Cu(52 nm)/Si (111).

-300 -200 -100 0 100 200 300-1.0

-0.5

0.0

0.5

1.0

Applied magnetic field (Oe)

Nor

mal

ized

Ker

r ro

tatio

n θ K

/θK

S

-300 -200 -100 0 100 200 300-1.0

-0.5

0.0

0.5

1.0

Nor

mal

ized

Ker

r ro

tatio

n θ K

/θK

S

Applied magnetic field (Oe)

-300 -200 -100 0 100 200 300-1.0

-0.5

0.0

0.5

1.0

Applied magnetic field (Oe)

Nor

mal

ized

Ker

r ro

tatio

n θ K

/θK

S

25

0 20 40 60 80 100 120 140 160 180

0

10

20

30

40

50

60

70

Ni/glass Ni/Si (111) Ni/Cu Ni/mica

Coe

rciv

e fie

ld (O

e)

Ni thickness (nm)

(a)

0 10 20 30 40 50 60 70 800

20

40

60

80

100

120

140

160

Ni/Cu/glass Ni/Cu/Si (111) Ni/Cu/mica Ni/Cu/Cu

(b)

Coe

rciv

e fie

ld H

c (O

e)

Ni thickness (nm)

Fig. 3. Coercive field HC vs Ni thickness for (a) Ni films on glass, Si, mica and Cu substrates

and (b) Ni/Cu/Substrates with substrates as indicated.

26

0 20 40 60 80 100 120 140 160 180

0.4

0.5

0.6

0.7

0.8

0.9

1.0Sq

uarn

ess

(θK

R/θ

KS)

(a)

Ni thickness t (nm)

Ni/glass Ni/Si (111) Ni/Cu Ni/mica

0 10 20 30 40 50 60 70 800.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

Ni/Cu/glass Ni/Cu/Si (111) Ni/Cu/Cu Ni/Cu/mica

(b)

Squa

rnes

s (θ

KR/θ

KS)

Ni thickness t (nm)

Fig. 4. Squareness S vs Ni thickness for (a) Ni films on glass, Si, mica and Cu substrates and

(b) Ni/Cu/Substrates with substrates as indicated.

27

0 20 40 60 80 100 120 140 160 180

40

60

80

100

120

140

160

180

200

Ni/glass Ni/Si (111) Ni/Cu Ni/mica

(a)Sa

tura

tion

field

(Oe)

Ni thickness (nm)

0 10 20 30 40 50 60 70 800

50

100

150

200

250

300

350

Ni/Cu/glass Ni/Cu/Si (111) Ni/Cu/Cu Ni/Cu/mica

Satu

ratio

n fie

ld (O

e)

(b)

Ni thickness (nm)

Fig. 5. Saturation field HS vs Ni thickness for (a) Ni films on glass, Si, mica and Cu

substrates and (b) Ni/Cu/Substrates with substrates as indicated.